Fig 3. The flow of information objects between STEP and SGML environments.
The goal of the architectural framework is to provide sharable data between the two environments without compromising either individual environment. This can only be achieved by providing a superset containing all of the concepts and functionality of both domains and where the data and operations are compatible and harmonized. The architecture must provide a functional and conceptual compatibility between the STEP constructs and the SGML constructs in order for interoperability to succeed between the two environments.
The architectural framework is shown in the following diagram as a set of flows between the SGML and the STEP environments.
Fig 3. The flow of
information objects between STEP and SGML environments.
The architecture presumes two environments: the STEP-compliant and the SGML-compliant. These are shown separately in the architecture as this will be the predominant situation. However, the architecture does not in any way preclude combining the environments in other ways.
The following steps form the work and control flow patterns of the architecture.
The request may be for any entity that is stored and is addressable (either by a unique global address or by a search query) within the STEP-compliant repository, e.g., an SGML_STRING, a tool-number, the pressure of a pump, etc.
The Standard Generalized Markup Language (SGML) specifies a standard way of defining document structures for the application of markup schemes based on generic coding, i.e., it provides rules for marking up text which enables the exchange of documents between different computer systems. SGML provides a consistent and precise manner of describing revisable documents for interchanging widely varying application data. It is an open-ended definition reflecting the diversity of information needs in industry, specifically trade and product definition data. SGML does not directly define any types of content data, and thus does not restrict the type of data contained in a document.
SGML is not a predefined set of tags that can be used to mark up documents. Neither is it a standardized template for producing particular types of documents. SGML is language for describing the component parts of a document that is to be transmitted from one computer to another. SGML is flexible enough to be able to describe any logically structured set of data, whether it be a form, memo, letter, report, book, encyclopaedia, dictionary or even a spreadsheet or database.
The SGML standard is based on the simple principle that documents are a combination of content and markup. Markup is represented by explicit start- and end-tags which indicate an element type, roughly the same as a STEP entity, a textual (or other) object such as a “title” or “paragraph”. Elements may also have associated attributes which refine, uniquely identify, or further qualify a particular instance of an element.
In an SGML representation of a document instance, markup and content are intermixed in one text stream. SGML parse software deconstructs the stream into a database, or builds a tree representation for editing software, or steps through the file as if it were a procedure in order to present it, e.g., print it out on paper.
SGML is a metalanguage. It is optimized for declaring structures that are meaningful for document information, SGML is built on the idea that a document text stream consists of three parts:
Whenever possible, SGML application developers have insisted on drawing a clear boundary between what a unit of text is and how it looks. Elements are marked up with tags indicating the element type. There are no element types corresponding to typographical commands, i.e., an element is a structural thing like a paragraph or a title, not a “bold”, “new page” or “Space down 3 lines” command.
The importance of this cannot be underestimated: Making this separation allows for the re-use without arbitrary limitation of all parts of an SGML document. This is a philosophy which has not always been practised within STEP, and the result is the inadvertent binding of information to one presentation and one opportunity for use. Inherent to T14's philosophy is the re-process ability of all content.
SGML is closely related to two other international standards: The Document Style Semantics and Specification Language (DSSSL) and the Hypermedia/Time-based Structuring Language (HyTime). DSSSL provides two standardized processes, the first is the Transformation Process which defines a process for the conversion of SGML document types, i.e., from one DTD to another DTD. The second is the Formatting Process which defines a standardized process for attaching style semantics to elements of an SGML document.
HyTime is a standard application of SGML which provides semantics for multimedia constructs, such as hyperlinking, synchronizing, etc. One of the features of HyTime provides a method of encoding standard queries, which can be in the HyTime query language HyQ or in other query languages such as SQL. With HyTime it is possible to encode multimedia applications for transfer between dissimilar platforms.
The ISO standard for Product Data Representation and Exchange (ISO 10303) aims are to provide the broadest possible range of industry with a neutral form for the capture and interchange of computerized models throughout the total life cycle of a product.
Contrary to SGML which allows for application defined tags to indicate the structure of data and leaves the semantics completely up to the interpretation of the application, STEP defines the structure and the semantics totally. A generic conceptual model provides a basis for the development of separate data models for specific application structures and semantics. The EXPRESS language is used to define application protocols (APs) using the generic concepts and their relationships.
The STEP architecture also includes an interchange file (Part 21) which is created by mapping from the generic resources using the appropriate AP.